Computerized Community Information Systems (CCISs, also known as Public Access Networks, Community Networks, and Free-Nets(1)) are becoming increasingly common in communities across the United States and Canada. In February 1996 there were more than 390 systems either operational or organizing. These systems focus on providing information that people in local communities need in their daily lives. To evaluate how well they accomplish their objectives, we must ask: (1) Who do the systems serve?; (2) What information needs do they try to meet?; and (3) How effectively do they meet those needs? To help answer these questions, we have developed a new taxonomy of information needs and services and have tested the taxonomy in a pilot study of four operational systems. The taxonomy consists of a three dimensional matrix, the axes of which are Situational (or Subject) Categories, Types of Help provided, and Socioeconomic Identifiers. The fourteen Situational Categories like education, governmental processes, and social services were further subdivided. Five Types of Help were identified: Advocacy, Counseling, Directional, Factual, and Interactive Communication. Socioeconomic Identifiers included variables like age group, educational level, gender and income. We tested the taxonomy by examining the several hundred services provided by four CCISs (Victoria Free-Net, Big Sky Telegraph, Blacksburg Electronic Village, and Mobile Free-Net), locating each service within the three-dimensional matrix. We found that: (1) Directional and factual help dominate the service offerings, but the types of information provided varied considerably from one system to another; (2) Services based on interactive communications were significant and growing; (3) Counseling services were present but limited; (4) Advocacy was almost non-existent; and (5) Information about social services was rare. Services oriented to middle income and upper middle income patrons dominated the four systems. Very few services were specifically targeted to low-income patrons. The pilot study indicated that the taxonomy is useful for evaluating CCIS services. It is particularly useful for identifying areas in which service offerings are weak or non-existent, thus it can be used to help a CCIS improve its services to better meet its objectives.